Gabapentin and Pregabalin for the Acute Post-operative Pain Management. A Systematic-narrative Review of the Recent Clinical Evidences

Transcription

1 716 Current Drug Targets, 2009, 10, Gabapentin and Pregabalin for the Acute Post-operative Pain Management. A Systematic-narrative Review of the Recent Clinical Evidences Mario Dauri*, Skerdilajd Faria, Antonello Gatti, Ludovica Celidonio, Roberta Carpenedo and Alessandro F Sabato Department of Anaesthesiology, Emergency and Intensive Care Medicine, University Hospital of Tor Vergata, Rome, Italy Abstract: Background: Gabapentin and pregabalin inhibit Ca 2+ currents via high-voltage-activated channels containing the 2-1 subunit, reducing neurotransmitter release and attenuating the postsynaptic excitability. They are antiepileptic drugs successfully used also for the chronic pain treatment. A large number of clinical trials indicate that gabapentin and pregabalin could be effective as postoperative analgesics. This systematic-narrative review aims to analyse the most recent evidences regarding the effect of gabapentinoids on postoperative pain treatment. Methods: Medline, The Cochrane Library, EMBASE and CINHAL were searched for recent ( ) randomized clinical trials (RCTs) of gabapentin-pregabalin for postoperative pain relief in adults. Quality of RCTs was evaluated according to Jadad method. Visual analogue scale (VAS), opioid consumption and side-effects (nausea, vomiting, dizziness and sedation) were considered the most important outcomes. Results: An overall of 22 gabapentin (1640 patients), 8 pregabalin (707 patients) RCTs and seven meta-analysis were involved in this review. Gabapentin provided better post-operative analgesia and rescue analgesics sparing than placebo in 6 of the 10 RCTs that administered only pre-emptive analgesia. Fourteen RCTs suggested that gabapentin did not reduce PONV when compared with placebo, clonidine or lornoxicam. Pregabalin provided better post-operative analgesia and rescue analgesics sparing than placebo in two of the three RCTs that evaluated the effects of pregabalin alone vs placebo. Four studies reported no pregabalin effects on preventing the PONV. Conclusion: Gabapentin and pregabalin reduce pain and opioid consumption after surgery in confront with placebo, but comparisons with other standard post-operative regimens are not sufficient. Gabapentin and pregabalin seem not to have any influence on the prevention of PONV. INTRODUCTION Gabapentin (GBP) [1-(aminomethyl)cyclohexaneacetic acid] an alkylated analogue of gammaaminobutyric acid (GABA) was introduced in 1993 in Europe and the following year in USA. It was first developed as an anticonvulsant drug and than the GBP potentials as an analgesic drug for the treatment of the neuropathic pain was described in mid 1990s [1, 2]. Pregabalin (PGL) [(S)-(+)-3- (aminomethyl)-- methylhexanoic acid] was indroduced in Europe and USA a decade after the GBP approval and has a similar pharmacological alkylated GABA analogue structure. Pharmacology Similar, in few pharmacological aspects but, different in others, both GBP and PGL produce several pharmacological *Address correspondence to this author at the Department of Anaesthesiology, Emergency and Intensive Care Medicine, University Hospital of Tor Vergata, Via Di S. Eufemia, 11, Rome, Italy; Tel: ; Fax: ; mario.dauri@fastwebnet.it The work should be attributed to the Department of Anaesthesiology, Emergency and Intensive Care Medicine, University Hospital of Tor Vergata, Rome, Italy. effects, including interaction with L-amino acid transporter (important for absorption from gastrointestinal tract and distribution across blood-brain barrier) [3, 4], inhibition of Ca 2+ currents via high-voltage-activated channels containing the 2-1 subunit, leading in turn to reduced neurotransmitter release and attenuation of postsynaptic excitability [- 10]. The decreased Ca 2+ influx reduces also the excitatory aminoacid [ex: glutamate] and substance P release, leading to suppression of neuronal excitability following nerve or tissue injury [11]. However, their antinociceptive effects occur primarily in the setting of neural sensitization after nerve or tissue injury and it seems to be minimal on normal physiological pain transmission [12]. A recent study in both laboratory animals and humans suggest an interaction between GBP and spinal alpha-2- adrenergic receptor systems in the rat and, furthermore, that GBP administration reduces cerebrospinal fluid norepinephrine levels in humans [13]. GBP exerts its analgesic action through a negative indirect interaction with the glycine binding site of NMDA receptors [14, 1]. SV2A, a ubiquitous synaptic vesicle glycoprotein that may prepare vesicles for fusion and serves as the target for levetiracetam and its analog brivaracetam (which is currently in late-stage clinical development) [16] /09 $ Bentham Science Publishers Ltd.

2 Gabapentin-pregabalin for Post-operative Pain Current Drug Targets, 2009, Vol. 10, No Table 1. Gabapentin and Pregabalin, Both Similar and Different on Mechanisms of Action, Pharmacokinetics, Interactions with Other Drugs, Uses, Dosage and Side Effects Gabapentin Pregabalin Year of approval Structure Mechanisms of action Pharmacokinetics Interactions Uses Dosage and administrati on Side effects 1993 in UK and Europe,1994 in USA 2004 in UK, Europe and USA 2[1(aminomethyl)cyclohexyl]acetic acid (GABA analogue) Interaction with L-amino acid transporter (important for absorption from gastrointestinal tract and distribution across blood-brain barrier) [3, 4] Activation of GABA B receptor (controverse) [17-19] Inhibition of Ca 2+ currents via high-voltage-activated channels containing the 2-1 subunit, reducing neurotransmitter release and attenuation of postsynaptic excitability [-10]. The decreased Ca 2+ influx reduces also the excitatory aminoacid (ex: glutamate) and substance P release, leading to suppression of neuronal excitability after nerve or tissue injury [11] and decrease AMPA receptor activation and noradrenaline release in the brain [97]. Interaction with spinal -2- adrenergic receptor systems in the rat, and GBP administration reduces cerebrospinal fluid norepinephrine levels in humans [13]. GBP exerts its analgesic action through a negative indirect interaction with the glycine binding site of NMDA receptors [14, 1]. SV2A, a ubiquitous synaptic vesicle glycoprotein that may prepare vesicles for fusion and serves as the target for levetiracetam and its analogue brivaracetam (which is currently in late-stage clinical development) [16]. K v7/kcnq/m - K + channels that mediate the M-current, which acts a brake on repetitive firing and burst generation [16]. Available only as oral preparation. Absorption dependent by a saturable L-amino acid transport: the bioavailability of GBP varies inversely with dose: 300 mg= 60%; 600 mg= 40%; 1600 mg=3% (steady state). Plasma peak = mg/l achieved h after ingestion of 300 mg (because of the dose-dependent absorption, plasma peak increases less than threefold when the dose is tripled) Volume of distribution: l/kg; cerebrospinal fluid concentration: 20% of plasma concentration; brain issue concentration: 80% the plasma level. No hepatic metabolism and it is eliminated unchanged in the urine with first order kinetic mechanism Elimination half-life: h. No microsomal enzyme induction. [90, 91, ] No pharmacokinetic interaction with anticonvulsivant drugs [90] Cimetidine decrease clearance of GBP (because decrease glomerular filtration) of 12% [90] Antacids reduce bioavailability of GBP when given until 2h post its administration [102]. General tonic-clonic seizure, partial seizures, peripheral neuropathic pain, diabetic peripheral neuropathy, post-herpetic neuralgia and acute pain [22]. To give 3 time/day because of the short half-life Epilepsy in adults 2400 mg/day; Epilepsy in children 2-3 mg/kg/day Neuropathic pain: mg/day Somnolence (1,2%), dizziness (10.9%), asthenia (6%), convulsions (0.9%), reversible acute renal allograft dysfunction and exacerbation of myasthenia gravis [22] More frequent nausea and vomiting than PGL [21] (S)-3-(aminomethyl)--methylhexanoic acid (GABA analogue) Interaction with L-amino acid transporter (important for absorption from g.i. tract and distribution across blood-brain barrier) [3, 4] No evidence of interaction with GABA B receptor [20] Inhibition of Ca 2+ currents via high-voltageactivated channels containing the 2-1 subunit, leading in turn to reduced neurotransmitter release and attenuation of postsynaptic excitability [-10] The decreased Ca 2+ influx reduces excitatory aminoacid (ex: glutamate) release leading to decreased AMPA receptor activation and noradrenaline relase in the brain [97]. Inhibitory modulation in neocortex, amygdale e hippocampus [21]. Available only as oral preparation. Absorption not saturable (linear pharmacokinetic profile) Plasma peak ( mg/l) reached in within 1 h. Average bioavailability> 90% independent of dose. Elimination half-life:.-6.7h (independent of dose). No hepatic metabolism and renal excreted (98% unchanged in urine, 0.9%Nmethylated derivative); elimination proportional to creatinine clearance. [92, 93, 101] No pharmacokinetic interactions Concurrent intake reduces peak plasma levels by 2 30% and increases the time to peak by 3 hours [92, 93, 101] Peripheral neuropatic pain, partial seizures, diabetic peripheral neuropathy, posterpetic neuralgia, general anxiety disorders, fibromyalgia and acute pain [21]. Neuropathic pain: mg thrice a day Acute pain: mg/day Fibromyalgia: mg/day Somnolence (22%), dizziness (29%), convulsions, weight gain, myoclonus, aterixis and gynecomastia [21] K v 7/KCNQ/M - K + channels that mediate the M-current, which acts a brake on repetitive firing and burst generation [16]. Activation of GABA B receptor by GBP is almost controversial [17-19], meanwhile no evidence of interaction with GABA B receptor is reported for PGL [20]. Inhibitory modulation of PGL is also exerted in neocortex, amygdale e hippocampus [21]. A summery of GBP and PGL similarities and differences on mechanisms of action, pharmacokinetics, interactions with other drugs, uses, dosage and side effects is reported in Table 1. Gabapentinoids Indication for Use These anticonvulsants have been used for the treatments of a wide variety of disorders including general tonic-clonic seizure, partial seizures, peripheral neuropathic pain, diabetic peripheral neuropathy, post-herpetic neuralgia and acute pain [22]. Furthermore, GBP and PGL seem to be effective on various forms of pruritus, including uraemic pruritus, intractable hiccups and hot flushes in post-menopausal women [23]. With respect to the GBP benefits on phantom

3 718 Current Drug Targets, 2009, Vol. 10, No. 8 Dauri et al limb pain the opinions are inconclusive. Bone et al. reported that GBP was efficacious to treat this syndrome [24] in discordance with other authors [2, 26]. GBP has been used also for attenuating haemodynamic response to tracheal intubation [27, 28] and on reducing postoperative delirium [28], and like an alternative to benzodiazepines in the treatment of alcohol withdrawal [29]. The first published PGL randomised clinical trials (RCTs) in 2001 [30], and the first published GBP RCTs in 2002 [31, 32] reported cheering results for the treatment of the post-operative pain. The following years, several authors reported their experiences regarding the use of GBP [33-68] and PGL [69-76] for the post-operative pain management. The main objective of this review is focusing the attention on the recent evidences for the analgesic properties and the adverse effects of GBP and PGL as treatment for the post-operative pain. METHODS The Selection of the Studies The articles research was performed using MEDLINE, Cochrane Controlled Trials Register (CCTR), EMBASE, and CINAHL databases. lists of the retrieved articles were also searched. The period of publication were established from 1990 to The first studies of PGL use after major surgery pain management were published in 2006 and furthermore previous systematic narrative [11, 28, 77, 78] or meta-analytic [79-8] reviews reported data from GBP clinical trials until 2006 (included). A total of three trials on PGL use for post-operative pain management [30, 69, 71] have been included in only one review [11] and the other three [77-79] included only one trial on postoperative dental pain published in 2001 [30]. Thus, we decided to take in consideration only recent (from ) RCTs which investigated the analgesic effects of GBP or PGL in adult patients (age range 18 years and above) underwent to surgical procedures. Trials were included if they were randomized, double-blind, active or placebo controlled, had at least 10 subjects per study group, and reported both analgesic consumption and pain scores. Trials studying both pre and post - operative GBP or PGL were included, also if these drugs were part of a multimodal technique. Works that reported information about use of GBP and PGL in different settings (treatment of epilepsy, neuropathic pain, and other type of pain or that produced by other drugs), nonrandomised or experimental pain studies, case reports, and clinical observations and editorials were excluded. No language restrictions were applied and no investigators were contacted. Abstracts published in congresses acts and unpublished studies were not considered. The following search terms were included: gabapentin; pregabalin; postoperative pain; postoperative analgesia; pain measurement; postoperative nausea and vomiting; postoperative outcome. The lists of the selected studies and reviews were checked for additional citations. The last search was performed on 30 April Literature Information, Outcome Measures The following information were collected: 1) publication details, 2) patient population, number of patients, age, sex (male/female), settings and surgical procedure, 3) study design, description of drugs administration and follow-up, ) intra- and postoperative analgesics and type of administration, 6) outcome measures, pain and analgesic consumption, 7) withdrawals and adverse effects. Bibliographic research was performed and data were collected independently by three investigators (S. F, L. C, R. C) and reviewed by the others (M. D, A. G., AF. S). In order to measure the likelyhood of bias in pain research reports, the calculation was performed for each of clinical trials included in the review (Table 2) [86]. RESULTS Post-surgical Pain At the time of writing, we found an overall of 0 original works that reported the gabapentinoids use in 4248 surgical patients [31-76, 87-89]. Nine of them studied the effects of PGL [30, 69-76] and the other 41 those of gabapentin [31-68, 87-89] in different surgical settings as gynaecological, abdominal, neurosurgery, musculoskeletal, thoracic, head, neck and breast. Gabapentin studies involved totally 3343 patients, 114 of them received gabapentin and were confronted with 282 patients who received combination of Table 2. Jadad Score Calculation (from 0 to ). Instrument to Measure the Likelihood of Bias in Pain Research Reports 1 2 Items (based on randomization, blinding, and dropout) Was the study described as randomized (this includes words such as randomly, random, and randomization)? Give 1 additional point if: the method to generate the sequence of randomization was described and it was appropriate (table of random numbers, computer generated, etc.) Deduct 1 point if: the method to generate the sequence of randomization was described and it was inappropriate (patients were allocated alternately, or according to date of birth, hospital number, etc.) Was the study described as double blind? Give 1 additional point if: the method of double blinding was described and it was appropriate (identical placebo, active placebo, dummy, etc.) Deduct 1 point if: the study was described as double blind but the method of blinding was inappropriate (e.g., comparison of tablet vs. injection with no double dummy). 3 Was there a description of withdrawals and dropouts? Score, Yes/No 1/0 1/0-1/0 1/0 1/0-1/0 1/0

4 Gabapentin-pregabalin for Post-operative Pain Current Drug Targets, 2009, Vol. 10, No GBP - gabapentin PGL - pregabalin RCTs - randomised clinical trials Fig. (1). Flow chart of the review. gabapentin and other analgesic, 389 patients who received other analgesics and with 118 patients receiving placebo as controls. PGL studies involved totally 90 patients, 323 of them received gabapentin and were confronted with 194 patients who received combination of PGL and other analgesic, 137 patients who received other analgesics and with 21 patients receiving placebo as controls. Furthermore, we found a total of 4 systematic narrative [11, 28, 77, 78] and 7 meta-analytic [79-8] reviews reporting information from RCTs on gabapentin and PGL use for the treatment of post-operative pain, published from 2004 to The flow chart in the Fig. (1) reports generally the selection of the studies included. This review focuses particularly the attention to the 2 gabapentin [47-68, 87-89] and 8 PGL [31-76] studies published from 2006 to 2009, which involved a total of 2668 patients and to the meta-analysis [79-8]. Gabapentin RCTs Three of the 2 gabapentin studies were not RCTs [87-89]. Nissman s et al. work was a prospective cohort study that included a total of 141 patients and reported information about analgesic properties of gabapentin after keratectomy [88], meanwhile Parsa et al. analyzed the gabapentin and celecoxib combination in aesthetic surgery (118 patients) [89]. In all these works the results were then compared with previous data as control patients concluding that gabapentin administration significantly reduces postoperative pain and opioid requirements. Van Elstraete, found that the median effective dose of pre-emptive gabapentin on postoperative morphine consumption after posterior lumbar spinal fusion was 21.7 mg kg -1 (9%CI: mg kg -1 ) [87]. More detailed information regarding each of 22 gabapentin RCTs (an overall of 1640 patients) included in the review is reported in Table 3. There has been tested a large modality of GBP administration for post-surgical analgesia. All the RCTs administered pre-emptive gabapentin or its combination with other analgesic drugs. Thirteen of all presented RCTs evaluated pre (10 RCTs) [0, 2, 4, 6, 8, 61, 63, 64, 67, 68] and both pre/post-operative (3 RCTs) [48, 49, 9] doses of GBP alone vs placebo patients. Eight RCTs studied gabapentin in confront with other analgesics (dexamethasone, lornoxicam, celecoxib, rofecoxib, acetaminophen, clonidine); five of them used only pre-emptive analgesia [3,, 7, 62, 6] and 3 considered both pre and post-operative mixture administration [47, 1, 66]. Only one work confronted different dosages of pre-emptive gabapentin, concluding that increasing the dose of gabapentin (300 to 1200 mg), appears to significantly decrease the severity of postoperative pain and total fentanyl consumption during the first 24 hours after myomectomy [60]. Variable GBP preemptive doses from 300 mg to 1600 mg were administered achieving the highest dosage of 3200 mg/day at the surgery day in one RCT [9]. No studies considered the comparison of pre-emptive and post-incisional or post-surgery GBP administration. The follow up period was no more than 24 h in thirteen studies [0, 3-8, 60-63, 67, 68]. In the other

5 720 Current Drug Targets, 2009, Vol. 10, No. 8 Dauri et al nine studies the patients observation varied from 2 to 7 POD [47-49, 1, 2, 9, 64-66], and four of them inter-viewed the patients by phone also one month [48, 9, 66] and three month [47] post-surgery. No studies established the optimal post surgical GBP treatment duration. Pain assessment has been performed using a Visual Analogue Scale (VAS) in 18 RCTs [48, 0, 2-67], meanwhile three studies used a 11- point [47, 49] or 4-point [1] Verbal Rating Scale (VRS). One study did not consider pain evaluation but only postoperative PCA fentanyl consump-tion, focusing the attention to the anti-emetic gabapentin effects [68]. One study assessed the pain only at rest [6], eleven RCTs evaluated pain both at rest and on movement [48, 0, 1, 3,, 8, 9, 61, 63, 64, 66] and the remaining 9 works did not specify whether pain was measured at rest or with movement [47, 49, 2, 4, 7, 60, 62, 6, 67]. The total rescue analgesics consumption has been used as another outcome for the assessment of the post-operative analgesia efficacy. In nine works the patients have been instructed to use an i.v. PCA pump with morphine [47, 48, 3,, 9, 62, 6] or fentanyl [64, 68] without continuous infusion. Seven studies reported i.v. morphine [0, 2, 6] or fentanyl [4, 60, 66, 67] boluses on demand administered by the personnel. Two RCTs used an epidural PCA pump with [61] or without continuous anaesthetic infusion [49]. Other rescue analgesics have been administered on demand in 13 studies that was alone [1, 7, 8, 63] or in adjunction to the previous i.v. opioid boluses [47, 48, 2, 4, 9, 64, 66] or epidural PCA [49, 61]. Furthermore, three RCTs have treated all the patients with a standard analgesic regimen as well as the on demand therapy [1, 2, 64]. Three studies [2, 4, 6] registered also the time elapsed from the end of the surgery to the first analgesic demand as a further outcome. Table 3. Gabapentin for the Postoperative Pain Management. Randomized Clinical Trials setting, Demographics, sample size, dosing, active/control, follow-up Anaesthesia, Intra-operative and postoperative analgesics Abdominal and pelvic surgery Turan, abdominal hysterectomy. 47 Bartholdy, 2006 laparoscopic sterilization 3 Durmus, 2007 abdominal hysterectomy: Sex: All females Age: A, 3±13; B, 0±11 C, 49±14; D, 1±11 Orally 1 h before surgery. Than, the same dosage at the 1 st and 2 nd POD. Group A (n=2): GBP 1.2 g + PL, Group B (n=2): rofecoxib 0 mg + PL. Group C (n=2): GBP 1.2 g + rofecoxib 0 mg Group D (n=2): PL Follow up: at 1, 4, 8, 12, 16, 20, 24, 30, 36, 42, 48, 60, and 72 h Sex: All females Age: A, 37 (28 4); B, 38 (27 4) Orally 30 min before surgery. Group A (n=38): GBP 1.2 g + lornoxicam 8 mg, Group B (n=38): lornoxicam 8 mg + PL Follow up: 2h and 4h. Sex: All females Age: A, 48±7; B, 49±6; C, 48±7 Orally 1 h before the anaesthesia Group A (n=2): GBP 1.2 g Group B (n=2): GBP 1.2 g + acetaminophen 20 mg kg -1 Group C (n=2): PL Follow up: 1, 2, 4, 6, 24 h General anesthesia. IV fentanyl, 2 g kg -1 and morphine, 2 mg. PCA IV with morphine 2 mg bolus, 10 min lockout interval. Oral acetaminophen 00 mg + codeine 30 mg on demand. General anaesthesia remifentanil 0.4 μg kg-1 min -1, alfentanil 0. mg i.v. post-operative: PCA morphine, mg initial bolus, supplemental bolus doses of 2. mg, 10 min lock-out. Anesthesia: general anesthesia fentanyl 1 g kg -1 at induction, mg morphine 1 min before the end of surgery PCA morphine 2 mg bolus, 1 min lock-out, 3 mg maximum 4 h limit. Lower pain scores at rest in groups A, B and C vs D at 4 h, 8 h, 16 h and 20 h (all p<0.0) and in groups A and C at 12 h (p<0.001) and 24 h (p<0.0). The pain with movement was lower in group A at 4 h (p<0.01), in groups A, B and C at 8 h (all p<0.0) and in groups A and B at 20 h (p<0.0) vs group D. Reduced morphine consumption in groups A, B and C at 1, 8, 24, and 30 h after surgery (all p<0.0). Total PCA morphine use was decreased by 24%, 43%, and 0% in groups A, B and C vs D. Shorter period of time of opioids requirement in group C vs D. Less oral analgesic consumption in groups B and C vs D. At the 72-h follow-up, all of the patients in group C were completely satisfied vs 32%, 64%, and 72% in groups D, B, and A. Less nausea in group C vs group D, (all p<0.0). Ambulation, hospitalization, and recovery s quality were not different. Thirty-two (84%) patients in group A and 37 (97%) patients in group B did require morphine, (p=0.049). No difference in the morphine consumption (group A, 10.±7.1 vs B 13.7±7.4 mg, p=0.06) and on VAS. VAS at rest: At 2h: group A 13 (7 28) vs B 22. (11 3), p=0.06; At 4h: group A, 4 (0 10) vs B, 6 (1 11), p=0.22 VAS on movement: At 2h: group A 8 (4. 26) vs B 20(6. 29), p= 0.26; At 4h: group A, 3 (0 10) vs B, (1 14) p=0.49. Side-effects were similar between groups. Higher morphine consumption in group C vs A and B (p<0.0), and higher morphine consumption in group A vs B (p<0.0) at all time points. Higher VAS scores at rest and at movement at all time points in group C vs A and B (p<0.0). SpO 2 at 24 h was lower in group C vs A and B (p<0.0). Lower sedation in group C vs A and B until 4 h (p<0.0). Higher sedation in group C vs A and B at 24 h, the difference was only statistically significant for group B (p<0.0). At all time points, the patient dissatisfaction scores were higher in group C vs groups A and B (p<0.0). No significant difference in terms of the side-effects.

6 Gabapentin-pregabalin for Post-operative Pain Current Drug Targets, 2009, Vol. 10, No (Table 3). contd.. setting, Fassoulaki 2006; abdominal hysterectomy 48 Fassoulaki, 2007, abdominal hysterectomy. 9 Demographics, sample size, dosing, active/control, follow-up Sex: all females Age: A, 42±.6; B, 42±6.2 orally, starting at pm the day before surgery, than every 6h until the th POD Group A (n=30): GBP 400 mg Group B (n=30): PL Follow up: at 2, 4, 8, 24, 48, 72, 96, 120 h, and 1 month by phone. Sex: All females Age: A, 40±7.3; B, 40±7.7 Drugs administration Group A (n=27): premedication with GBP 1.6 g, than GBP 400 mg every 6 h, starting at p.m. the day before surgery till 7 th POD Group B (n=24): PL as group A. Follow up: at 2, 4, 8, 24 h, the 2 th - 7 th POD. Than 1 month by phone. Anaesthesia, Intra-operative and postoperative analgesics general anesthesia fentanyl μg kg -1 at induction. No other analgesics were described PCA with morphine 1mg ml -1, bolus 1mg, lock out of 7 min; acetaminophen 300 mg+ codeine 30 mg General anaesthesia. morphine 0.1 mg kg -1, acetaminophen 1.2 g. Continuous wound infusion with ropivacaine 0.7% (group A) or normal saline (group B) at 2 ml h -1. PCA morphine 1mg boluses, 7 min lockout. Oral acetaminophen - codeine on demand. No differences in morphine consumption at each time point and totally in the first 48h (p=0.09). At 2h: A, 9.0±4.3; B, 9.7±3.. At 4h: A, 10.±4.4; B, 13.4±.1. At 8h: A, 13.±.; B, 16.9±6.3. At 24h: A, 20.3±7.9; B, 2.7±11.2. At 48h: A, 28.4±12.1; B, 33.0±1.7. No difference on tablets of acetaminophen-codeine consumption (p=0.42). At 72h: A, 1.0 (0-4); B, 1.0 (0-4). At 96h: A, 1.0 (0-3); B: 0. (0-2). At 120h: A, 0.0 (0-2); B: 0.0 (0-3). No differences on pain at rest and after cough (p=0.46, and p=0.34 respectively). Less painful patients, (OR=0.16; 9% CI, ) and lower pain intensity OR=0.36; 9% CI, ) 1 month after surgery at group A vs B, p= Groups were similar regarding: VAS values at rest and after cough, p=ns, nausea or vomiting, dizziness, and sedation during all the point time follow up. Number of patients who required analgesics at home the first month after surgery did not differ between the two groups, P=NS. Overall cumulative morphine consumption at 48h, group A: 31.6±13.2 mg vs group B: 0.6±20. mg, p< During the postoperative days 3 7, the group A consumed fewer acetaminophen-codeine than group B, p= Fewer patients experienced pain 1 month after surgery in group A vs group B, P=0.04). Said-Ahmed 2007 myomectomy 2 60 Gilron, 2009 ambulatory laparoscopic cholecystectomy. 66 Sex: all female Age: A, 3 ± 8; B, 39 ± ; C, 38 ± 6; D, 36 ± 7 orally, 2h before surgery, Group A (n=20): GBP 300 mg Group B (n=20): GBP 600 mg Group C (n=20): GBP 1.2 g Group D (n=20): PL Follow up: 2, 6, 12, 24 h Sex, M/F: A: 6/24; B: 11/19; C: 8/21 Age: A, 41. (21 81); B, 49. (30 77); C, 46 (24 81) Group A (n=30): oral meloxicam 1 mg 1 h preoperatively, than the POD 1 and 2 Group B (n=30): oral GBP 1.2 g before surgery, 400 mg on the evening of surgery and GBP 400 mg x 3 on POD 1 and 2 Group C (n=29): oral meloxicam 1 mg/day for 3 days+gbp 1600 mg the day of surgery and GBP 1200 mg daily on POD 1 and 2. Follow up: every 30 min until discharge, at POD 1, 2, and 30. general anesthesia fentanyl 2 μg kg -1 at induction; no other analgesic were described fentanyl 2 μg kg -1 on demand general anaesthesia and trocar insertion sites were infiltrated with up to 1 ml of 0.2% bupivacaine. IV fentanyl 2 g kg -1 IV fentanyl g every 3 min as needed. Upon discharge from hospital, patients were prescribed either codeine mg or morphine 10 mg PO every 3 h as needed. With increasing the dose of GBP (300 to 1200 mg) there was a reduction in VAS scores compared to PL at all time points, which reached significance for group C. 2h: A, 4.±1.9; B, 4.4±2.1; C, 3.0±1.4; D,.0±1. (p= 0.004). 6h: A, 3.1±1.2; B, 2.8±1.3; C, 1.9±1.2; D, 4.2±1.1, (p<0.001). 12h: A, 2.±1.1; B, 2.2±1.1; C, 1.±0.9; D, 3.3±1.1 (p<0.001). 24h: A, 1.9±1.1; B, 1.7±0.7; C, 1.3±0.; D, 2.±1.2, (p=0.014). Significantly lower fentanyl consumption in group C ( g): A, 270±90; B, 20±8; C, 190±80; D, 340±9 (p<0.001). No difference in side effects (p=ns) Lower rest pain at 60 min in the group B vs A (p=0.003). Pain decreased from 60 to 120 min (p=0.000, 0.00, < for pain evoked by peak expiration, sitting, and cough, respectively), but no differences between groups (p=0.7, 0.3, 0.3, and 0.08 for shoulder pain, pain evoked by peak expiration, sitting, and cough pain, respectively). Lower cough pain at 60 min in the group B vs A (p=0.01). On POD 1, 2, and 30, no differences between groups on pain. No differences among groups in opioid consumption until POD 3. No significant effect of treatment by time interaction for any of the spirometric measures (p=0.44, 0.07, and 0.3 for PEF, FEV1, and FVC, respectively). All spirometric measures improved from 60 to 120 min (p<0.0003), but groups were not different for PEF, FEV1, and FVC (p=0.3, 0.9, 0.9, respectively). PEF was higher in group C vs A at 120 min (p=0.02). Less nausea in the group C vs A (p=0.016) but not vs B (p=0.8). Median time (hours) from PACU admission to meeting PACU discharge criteria was 2.92, 2.83, and 2.7 for group A, B and C, respectively (p=ns). Mean time (days) from surgery to return to work was 13.6, 11.7, and 10.6 for groups A, B, and C respectively (p=ns).

7 722 Current Drug Targets, 2009, Vol. 10, No. 8 Dauri et al (Table 3). contd.. setting, Demographics, sample size, dosing, active/control, follow-up Anaesthesia, Intra-operative and postoperative analgesics Pandey, 2006, laparoscopic cholecystectomy 68 Sex, M/F: A, 2/100; B: 13/107 Age: A, 42.8±11.4; B, 41.8±11.1 orally, 2h before surgery, Group A (n=12): GBP 600 mg Group B (n=12): PL Follow up: 2, 6, 12, 24 h. general anesthesia induction with fentanyl 3 g kg -1 ; manteinance not described fentanyl (dosage not described) Significantly higher fentanyl consumption in group B vs A. A: 221.2±92.40 g ; B: 0.9±82.0 g (p=0.01) The incidence of PONV during the first 24 hr was significantly lower in A group (37.8%) than B (60%), (p =0.04). No difference in the PONV severity. A: mild 9 patients, moderate 31 patients, severe 6 patients. B: mild 10 patients, moderate 2 patients, severe 13 patients. Similar incidence of side effects in both groups. Koç, 2007 varicocele surgery 4 7 Sex: All males Age: A, 39.±19.3; B, 38.4±17.4; C, 3.3±18.0; D, 41.1± h before surgery Group A (n=20): oral 800 mg of GBP + IV 2 ml saline Group B (n=20): oral PL + IV 8 mg dexamethasone Group C (n=20): 800 mg GBP+8 mg dexamethasone Group D (n=20): oral PL + IV 2 ml saline Follow up: at 1, 2, 4, 6, 12, and 24 h. general anesthesia remifentanil 0. g kg -1. min -1 at induction then it was reduced to 0.2 g kg -1. min -1. When VAS>3, tenoxicam 20 mg IM was administered Lower HR and MAP in group C at 1, 3,, and 10 min after intubation vs A and B (p<0.0) and D (p<0.001). Hemodynamics were similar in group A and B, but lower than group D (p<0.0). Less remifentanil consumption in group C (249.1±8.7 mg) vs A (408.±139.7 mg) and B (409.2±136.6 mg) (p<0.0) and D (74.7±119.7 mg) (p<0.001). Values in group D were higher than in group A and B (p<0.0) but A and B were similar. Similar MAP and HR among groups at each time point. Less pain in group C at 30 min, 1, 2, 4, 6, and 12 h vs A, B (p<0.0) and D (p<0.001). Values in group D were higher than in group A and B (p<0.0), but group A and B were similar. Lower tenoxicam consumption in group C (0 mg) vs A (80 mg) and B (80 mg) (p<0.0) and vs D (300 mg) (p<0.001). Less PONV in group C vs the other groups (p<0.001).ponv in group A and B were similar, but less than in D (p<0.0). No differences in other side effects. Significant differences in median VAS score at all time points (p<0.0) At 0 h: A: VAS 1 (0-2), B: VAS 2 (1-3) At 1 h: A, VAS 3 (1-3); vs B, VAS 3 (2-) At 2 h: A, VAS 3 (2-3); vs B, VAS 3 (3-) One patient (0.02%) in A group and 10 patients (28%) in B group required additional IV analgesic (p=0.012). Two patients in A group and 9 patients in B group had nausea (p=0.022). None patients in A group and 4 pz in B group had vomiting (p=0.114). No differences in side effects. Mohammadi, 2008 laparoscopic surgery for reproductive technologies 3 67 Sex: all female Age: A, 31.3±.4; B, 31.9±.6 Orally 1h before surgery Group A (n=3): GBP 300 mg Group B (n=3): PL Follow up: 0, 1, 2h. general anaesthesia induction wit fentanyl 2 g kg -1 ; manteinance not described Fentanyl as rescue analgesic (dose not described) Mohammadi, 2008 Abdominal (gynaecological/ge neral surgery) 4 62 Sex, M/F: A, 24/16; B, 23/17; C: 2/1 Age: A, 39±12; B, 3±13 C, 40 ±12 orally 1 h before surgery Group A (n=40): 300 mg GBP Group B (n=40): 0.2 mg clonidine Group C (n=40): PL Follow up: at 0, 1 and 6 h general anesthesia fentanyl 3 μg kg -1 for induction; fentanyl 1 μg kg -1 h -1 for maintenance. In the PACU: iv morphine titrated 2 mg every 10 min in order to obtain VAS<3; PCA: bolus 1 mg lockout interval 10 min. VAS score>3 significantly more frequent in B e C group than in A. PACU: A, 2%; B, 13%; C, 29%, (p=0.001). At 1 h: A, 19%; B, 36%; C, 29%, (p=0.001). At 6 h: A, 33%; B, 37%; C, 39%, (p=0.027) Morphine consumption at the PACU: A, 1.6±1. mg; B, 1.9±.mg; C, 4.7±7. mg (A vs B p=0.04; A vs C p=0.024; B vs C p=0.032). Morphine consumption during the first 6 h: A, 12.1±19.9 mg; B, 13.1±12.6 mg; C, 18.0±1.8 mg (A vs B, p=0.07; A vs C, p=0.023; B vs C, p=0.02. Side effects were not different between the groups. Lower VAS scores in groups A e B vs group C at 1, 12, 24 and 48h (p<0.0). At 1h: A, 4.24±0.4; vs B, 4.48±0.8; vs C, 6.39±0.48, (A and B vs C, p<0.0). At 4h: A, 4.2±0.3; vs B, 4.62±0.44; vs C,.81±0.40, (A vs C, p<0.0); At 8h: A, 3.1±0.31; vs B, 4.86±0.41; vs C, 6.10±0.47, (A vs B and C, p<0.0). At 12h: A, 2.92±0.32, vs B3.43±0.38, vs C, 4.94±0.40 (A and B vs C, p<0.0). At 24h: A, 1.81±0.30; vs B, 1.76±0.30; C, 3.48±0.40 (A and B vs C, p<0.0). At 48h: A, 0.64±0.19; vs B, 1.12±0.28; vs C, 2.17±0.38 (A and B vs C, p<0.0). Lower morphine consumption in groups A and B vs C till 24 h after surgery (p<0.0); no difference at 48 h. No differences in side effects. Ghafari 2009; abdominal hysterectomy 2 6 Sex: all female Age: A, 4±1; B: 44±1; C: 44±1 at 10:00 pm the night before and 1 h before surgery Group A (n=33): 300 mg GBP Group B (n=33): 100 g clonidine Group C (n=33): PL Follow up: 1, 4, 8, 12, 24, and 48 h. general anesthesia induction with fentanyl 2. g kg - 1 ; maintenance with fentanyl 1 g kg -1 every 30 min PCA with morphine 1 mg ml -1, bolus 1 mg, 7 min lock out period.

8 Gabapentin-pregabalin for Post-operative Pain Current Drug Targets, 2009, Vol. 10, No (Table 3). contd.. setting, Al-Mujadi, 2006 thyroid surgery. 0 Mikkelsen, 2006 tonsillectomy in adults 1 Demographics, sample size, dosing, active/control, follow-up Sex: M/F A: 9/26; B: 10/27 Age: A: 4 ± 13; B: 49 ± 1 Orally 2 h before surgery. Group A (n=37): GBP 1.2 g Group B (n=3): PL Follow up: At 2, 6, 12, 18 and 24 h. Sex, M/F: A, 9/14; B, 7/21 Age: A, 31 (18 43); B, 27. (18 3) Drug s administration: Group A (n=23): Orally 1 h before surgery GBP 1.2 g+rofecoxib 0 mg, than GBP 600 mg x 2 on the day of operation and GBP 600 mg x 3 for the next days. Group B (n=28): rofecoxib 0 mg before anesthesia and PL as group A. Follow up: at 2, 4 h and at 1 POD Anaesthesia, Intra-operative and postoperative analgesics Head, neck, thoracic and breast surgery Anesthesia: General anesthesia fentanyl 2 3 g kg -1 i.v. at the induction. Morphine 3 mg iv bolus doses were given every five minutes until VAS pain scores were 4 or less at rest, and 6 or less with swallowing. General anaesthesia Sufentanil or alfentanil Postoperative: rofecoxib 0 mg daily. Ketobemidone 2. mg as needed. In the PACU, from 0 to 4 h post-operatively, iv morphine in incremental doses of 2. mg on request VAS at rest, at 0 h: group A 3.3±2.8 vs B 4.±0.9, p<0.01; at 2 h: A, 1.81±1.1 B, 4.4±1.9 p<0.001; at 6 h: group A, 1.40±0.7 vs B, 2.41±1.3, p<0.01; at 12 h: group A, 1.60±1.3 vs B, 2.60±1.6, p<0.01; at 18 h: group A, 1.1±0.7 vs B 2.±1.8, p<0.01; at 24 h: group A, 1.8 ± 1.6 vs B, 2.3±1.3 p<0.01.vas during swallowing, at 0 h: group A, 4.±1.3 vs B,.1±1.8, p<0.01; at 2 h: group A, 2.6±1.6 vs B,.0±1.7, p<0.001; at 6 h: group A, 2.3±1.2 vs B, 3.13±1., p<0.01; at 12 h: group A, 2.2±1.2 vs B, 3.8±1., p<0.01; at 18 h: group A, 2.±1.4 vs B, 3.6±1.9, p<0.01; at 24 h: group A, 2.3±1.2 vs B, 3.±1.1, p<0.01. Les morphine consumption in group A (1.2 ± 7.6 mg) vs B (29. ± 9.9 mg), (P < 0.001). No differences on PONV. No statistically significant difference between the groups regarding: pain scores at rest and during swallowing of 0 ml of water, awakenings caused by pain, sedation and nausea, at any time period. Reduced ketobemidone consumption in the first 24 h in group A: 2.0±2.0 vs B: 4.±3.0 mg, p = Three-fold more dizziness in group A vs B, p = and gait disturbance four-fold more frequently in group A, p=0.02, five-fold more vomiting in group A vs B (p=0.046). Brogly, 2008 thyroidectomy 3 63 Jeon, 2008 tonsillectomy 3 64 Huot, 2008 postthoracotomy shoulder pain 4 61 Sex, M/F: A, 3/19; B, 3/18 Age: A, 49 (18 63); B: 49 (2 72) Orally 2h before surgery. Group A (n=22): GBP 1.2 g Group B (n=21): PL Follow up: 1h, 3h, 6h, 9h, 12h, 18h, 24h. Sex, M/F: A, 18/14; B, 9/17 Age: A, 27.7±11.; B, 24.2±6.3 Orally, the night before and 1 h before surgery, Group A (n=32): GBP 600 mg Group B (n=26): PL Follow up: at 1, 2, 4, 8, 12, 24, 36 and 48 h, then for 7 days after discharge. Sex M/F: A, 11/12; B, 17/11 Age: A, 60.1±13.6; B, 60.0±8.7 Orally 2 hours before surgical incision. Group A (n=23): GBP 1.2 g Group B (n=28): PL Follow up: At 0, 4, 8, 12, 16, 20 and 24 h. General anaesthesia and superficial cervical plexus block IV sufentanil g kg -1 at induction than boluses of - 10 g IV acetaminophen 1 g or 0 mg IV tramadol. General anesthesia No analgesics were reported. PCA with 1% fentanyl 2 ml bolus, 10 min lockout time, dicolfenac sodium 7 mg i.m. as needed. Acetaminophen 32 mg and tramadol 37. mg daily for 9 POD. General anesthesia and thoracic epidural block. fentanyl 2 - μg kg -1 or sufentanil μg kg -1 iv at induction and further boluses as needed. Epidural bupivacaine 0.1% + fentanyl 2 μg ml -1-1 at 0.1 ml kg -1 hr Epidural bupivacaine 0.1% + fentanyl 2 μg ml -1 at 0.1 ml kg - 1 hr -1 + bolus of 0.1 ml kg 1 of epidural solution as needed. S.c. hydromorphone 1 2 mg. The total (median, range) analgesic consumption (paracetamol and tramadol) was 3 (0 ) in group A vs 3 (1 ) in B, p=ns. Tramadol was required in 27.3% patients in group A vs 23.% in B (p=ns). No significant differences between groups for VAS at rest and during swallowing. After 6 mo, 8 patients presented pain scores < 3 vs 2 patients in the preoperative period (p=0.04). It was significantly lower in group A. There was a trend toward greater burning sensation and numbness in group B, but p=ns. Fentanyl consumption: group A, 28.1±31. ml vs B, 9.7±41. ml, p= Diclofenac injections: group A, 0.1±0.3 vs B, 0.8±0.9, p= The pain score at rest (rvas) was highest at 2 h (group A: 4.1±2.2, group B: 4.±2.4) and lowest on 9 th POD (group A: 2.1±1.3; B: 2.±2.2) (p always NS). The swallowing VAS (svas) in group B was highest at 1 h: 6.8±2.6 2 and 2h: 6.8±2.4, and lowest to 3.6±2.1 on the 9 th POD. In the group A, the svas was highest at 8 h (.8±2.0) and lowest on the 9 th POD (3.4±1.). Lower svas in the group A than B only at 2 and 4 h (p= 0.04; p=0.04). No significant differences regarding patients satisfaction, drowsiness, PONV and headache. Similar amount of epidural solution in both groups (group A: ± 7.8 ml vs group B: ± 94.8 ml, p=0.06) and hydromorphone consumption (group A: 2.36 ± 2. mg vs group B: 2.6 ± 3.2 mg, p = 0.36). No differences on pain intensity at surgical site and at shoulder between groups, for the entire postoperative period. Twenty-three patients (82%) in Group B experienced shoulder pain vs 21 (91%) in Group A. Nausea, vomiting, and pruritus were similar in the two groups. Seven patients in Group A (30%) and eight patients in Group B (29%) received antiemetic treatment. At four hours, the incidence of sedation scores > 1 was greater in Group A (21/23 patients), vs Group B (18/28 patients; p=0.02). At 24 hr, /18 patients in Group B had sedation scores > 1, vs 0/28 patients in Group A (P = 0.0).

9 724 Current Drug Targets, 2009, Vol. 10, No. 8 Dauri et al (Table 3). contd.. setting, Demographics, sample size, dosing, active/control, follow-up Anaesthesia, Intra-operative and postoperative analgesics Turan, 2006, postoperative epidural analgesia after lower limb surgery 49 Adam, 2006 arthroscopic shoulder surgery. 2 Turan, 2007 tourniquet pain and I.V. regional anesthesia. 4 Montazeri, 2007 lower extremity orthopaedic surgery. 4 6 Prabhakar, 2007 Surgical brachial plexus exploration for injury 4 8 Sex: not reported Age: A, 4 (28 74); B, 0 (2 68) Drug s administration: Orally 1 h before surgery, than same drugs at 09:00 on the 1 st and 2 nd POD Group A (n=20): GBP 1.2 g. Group B (n=20): PL Follow up: at 1, 4, 8, 12, 16, 20, 24, 30, 36, 42, 48, 60, and 72 h. Sex, M/F: A, 18/9; B: 18/8 Age: A, 43±18; B, 47±1 Orally 2 h before surgery Group A (n=27): GBP 800 mg Group B (n=26): PL Follow up: two times the 1 st day and four times the 2 nd day after hospital discharge. Sex, M/F: A: 1/; B: 14/6 Age: A, 3±12; B, 39±14 Orally 1 h before anesthesia Group A (n=20): GBP 1.2 g Group B (n=20): PL Follow up: at 1, 2, 4, 6, 12, and 24 h. Sex, M/F: A, 26/9; B, 28/7 Age: A, 34.7±18.1; B, 34.6±17.8 Orally two hours before induction of anaesthesia. Group A (n=3): GBP 300 mg Group B (n=3): PL Follow up: at 2, 4, 12, and 24 h. Sex, M/F: A, 10/0; B, 9/1 Age: A, 27. (18-33); B, 31 (20-3) Orally 2 hours before surgery. Group A (n=10): GBP 800 mg Group B (n=10): PL Follow up: Every 1 h for 24 h. For the analysis, the mean values were taken at intervals of 0-6, 6-12 h, h, and h. musculoskeletal surgery General anaesthesia Fentanyl, 2 g kg -1 i.v. Epidural bupivacaine 0.12% ml, with fentanyl 1 g ml -1. epidural PCA with bupivacaine 0.12% and fentanyl 1 g ml -1, ml bolus, 10 min lockout, or oral 00 mg acetaminophen. brachial plexus block and general anesthesia intra-operative: remifentanil 1 g kg -1 at induction. post-operative: ketoprofene 10 mg x 2 and 400 mg acetaminophen + 30 mg dextropropoxyphene as needed. I.V. regional anaesthesia with lidocaine, 3 mg kg -1. Fentanyl 0. g/kg IV boluses if tourniquet pain score > 4 Diclofenac, 7 mg i.m. as needed general anesthesia fentanyl 2 μg kg -1 at induction and morphine 0.1 mg kg -1 before the start of the surgery morphine 0.0 mg kg -1 IV on demand. General anesthesia fentanyl 2 μg kg -1 at induction, than fentanyl boluses as needed. ketorolac iv as demanded by the patient or if VAS score was >0. Greater VRS pain scores at 1, 4, 8, 12, and 16 h after operation in group B (P<0.001). AUC for the pain scores showed a statistically significant difference for the first 24 h (P<0.001) but not at 72 h (P=0.7). Less VAS scores at group A until 20 h (P<0.001). Reduced PCEA requirements in the group A at 24, 48, and 72 h. Duration of PCEA usage was shorter in group A (7±9 vs 38±11, p<0.0); less oral analgesic consumption in the group A. (30±400 vs 700±23 mg, p<0.0). Groups were similar in times to the return of bowel function, resumption of dietary intake, and length of hospitalization. Fewer group A patients had motor block (P<0.0). Patient of group A were more satisfied. More dizziness in group A, p<0.0. No statistically significant differences in the VAS score. Cumulative use of supplement analgesics was similar in each group (group A: 6±3, B: 7±3 tablets). Time to the use of the first analgesic tablet was comparable in the two groups (group A: 768 ± 218 min vs B: 719±199 min). Side effects were comparable in both groups, except that headaches were more frequent in the group B (P=0.034). About 2% 30% of the patients considered their analgesia insufficient. However, scores for overall satisfaction were high (about 80 mm in both groups) and did not differ between the two groups. Reduced VAS scores for tourniquet pain in group A at 30, 40, 0, and 60 min after tourniquet inflation. Prolonged time to intraoperative fentanyl rescue (3±10 min vs 21±13 min, P <0.0), and total fentanyl requirement during surgery in group A (83±73 vs 3±47 g, p<0.0). Higher quality of anesthesia reported by the anaesthesiologist (4 [3 4] vs 2 [1 2]) and the surgeon (3 [3 3] vs 2 ([2 3]) in group A (P<0.0). Prolonged time to first postoperative analgesic request in group A (13±2 min vs 8±19 min, P<0.0). Lower VAS at 1 and 2 h after surgery (p<0.01) and decreased diclofenac consumption in group A (30±38 mg vs 60±63 mg, P<0.0). No differences regarding to adverse effects. Lower VAS scores in group A vs B at 2 h: group A,.±1.8 vs B, 72.3±14.0 at 4 h: group A, 7.3±19.3 vs B, 70.±18.1 at 12 h: group A, 4.7±16.0 vs B, 62.0± 23.3 at 24 h: group A, 44.6±17.6 vs B, 66.±2.7. All p<0.0. Less required morphine in group A, 1.4±2. mg vs group B 17.9±3.0 mg, p<0.0. Significant difference between the two groups in the first time of patients morphine demand after surgery (group A 31.6 ±1.9 min vs group B 26.7±7.1 min, p<0.0). No significant difference in the recovery duration between the two groups. Postoperatively, the adverse effects were similar between groups. Total intraoperative fentanyl requirement, group A: 200 (100-22) μg vs B: 237. ( ) μg, p= Intraoperative and postoperative hemodynamics was similar in the 2 groups. The VAS scores at rest: at 0h: A, 31.±11.6 vs B, 46±14.3, p=0.01. At 6 h: A, 37.±11.1 vs B, 47.±14.4, p=0.01. At 12 h: A, 38±10.3 vs B, 49±17.3, p=0.01. At 18 h: A, 34.± 4.4 vs B, 49±14., p= At 24 h: A, 38.±10.0 vs B, 4±13., p= The VAS scores during movement: At 0h: A, 2±14.8 vs B, 67±1.7, p=0.04. At 6 h: A, 2.±10.9, vs B, 6±19.6, p=0.04. At 12 h: A, 4±11.0 vs B, 66±17.8, p=0.04. At 18 h: A, 0.±10.1 vs B, 63±12., p=0.04. At 24 h: A, 4.±10.1 vs B, 66.±16.3, p=0.04. Rescue analgesic doses in group B, 2. (0-4) vs A, 0 (0-3), (p=0.004). No side effects requiring intervention were noted in both groups. GBP gabapentin; PL placebo; VAS Visual Analogue Scale; I.V. intravenous; POD post-operative day; PCA patient controlled analgesia; PCEA patient controlled epidural analgesia; HR heart rate; MAP - mean arterial pressure; AUC- area under the curve GBP provided better post-operative analgesia and rescue analgesics sparing than placebo in 6 of the 10 RCTs that administered only pre-emptive analgesia [0, 4, 6, 8, 64, 67, 68]. Three studies reported no GBP effects on pain

10 Gabapentin-pregabalin for Post-operative Pain Current Drug Targets, 2009, Vol. 10, No scores or rescue analgesics sparing [2, 61, 63]. One study that did not assessed pain reported only opioid sparing effect of GBP [68]. Of the RCTs that measured pain with movement, 3 demonstrated significantly reduced movementrelated pain in GBP patients [0, 8, 64]. Three RCTs compared pre and post-operative GBP administration with placebo. Turan at al found perioperative GBP treatment useful for pain reduction and epidural PCA solution sparing and oral acetaminophen reduction [49]. Fassoulaki et al. did not demonstrate any benefits in early pain management and analgesics sparing with perioperative gabapentin after hysterectomy [48]. The same authors reported opioid sparing but not pain reduction, in the same setting the following year [9]. However, in both works they found a significant pain reduction one month after surgery [48, 9]. Several RCTs compared GBP with other analgesics for post-operative pain management. In adjunction to dexamethasone [7] or to a selective COX-2 inhibitor as rofecoxib [47, 1] or lornoxicam [3] GBP did not seem to offer further benefits on pain management in comparison with GBP, dexamethasone, rofecoxib or lornoxicam alone. However, there were reported reduced intra [7] and post surgery [1, 7] analgesics consumption or number of patients requesting analgesics [3]. Meanwhile, Gilron et al. did not find substantial positive effects of GBP-meloxicam mixture with those of GBP and meloxicam alone on pain control and opioid sparing after laparoscopic colecystectomy [66]. Mohammadi et al. described positive GBP effects on reduction of pain and PCA morphine consumption when compared with clonidine after abdominal and pelvic surgery [62], but their assertions where not confirmed by Ghafari et al. after laparoscopic gynaecological surgery [6]. GBPacetaminophen combination provides better pain control and reduced PCA-morphine consumption in comparison with GBP alone or placebo after abdominal hysterectomy []. Different dosages of gabapentin have been confronted only by Said-Ahmed that reported increasing the pre-emptive dose of gabapentin (300 to 1200 mg), significantly decrease the severity of post-myomectomy pain and total opioid consumption during the first 24 hours after surgery [60]. Two of the 3 trials that evaluated the time elapsed from the end of surgery to the first analgesic request, reported that it was significantly longer in GBP group vs placebo [4, 6] but the other one did not find difference [2]. Related to a specific antiemetic effect, 14 RCTs suggested that GBP did not reduce nausea and vomiting following surgery when compared with placebo [49, 0, 2, 4, 6, 8, 9, 61, 63, 64, 67]. clonidine [62, 6], or lornoxicam [3]. Three RCTs that studied the gabapentin alone [68] or its combination with dexamethasone [7] or rofecoxib [47] reported a significant reduction of PONV in gabapentin patients vs placebo [47, 7, 68]. Furthermore, GBP-dexamethasone combination seems to have a synergic positive effect on PONV reduction in confront with GBP or dexamethasone alone [7]. Combined with COX-2 inhibitors, GBP seems to have similar incidence of PONV where confronted with GBP [47, 66] and rofecoxib [47] alone and less if confronted with meloxicam alone [66]. These findings are in contradiction with those of Mikkelsen at al who reported five-fold more incidence of PONV in GBP-rofecoxib group than rofecoxib alone group [1]. Gilron et al. demonstrated that there are not differences on lung function during treatment with either GBP, meloxicam or both after laparoscopic cholecystectomy as assessed by peak expiratory flow rate [66]. This findings are in contradiction with a previous work of the same authors where these improvements were enhanced even further when GBP was combined with another COX-2 inhibitor for abdominal hysterectomy [41]. Furthermore, they were not able to demonstrate benefits of gabapentin on PACU discharge and return to work [66]. The most frequent other adverse effects were sedation, dizziness, headache. No statistical differences regarding the adverse effects have been observed between GBP, other analgesics, combination or placebo in 17 RCTs [47, 0, 3, -68]. Two studies reported less headache in gabapentin group [2, 4] and other two trials found more dizziness with GBP administration [49, 1]. Meanwhile, Fassoulaki et al. excluded the patients having PONV or other side effects from the study [48]. Pregabalin RCTs More detailed information regarding each of the eight PGL RCTs (an overall of 707 patients) included in the review is reported in Table 4. There has been tested a large modality of PGL administration for post-surgical analgesia. All the RCTs administered pre-emptive PGL or its combination with other analgesic drugs. Three RCTs evaluated pre [70, 7] and both pre/post-operative [73] PGL alone vs placebo patients. The PGL other analgesic [dexamethasone [74], ibuprofen [72], celecoxib [69], and both acetaminophen-dexamethasone [76], combination has been studied by four RCTs, three of them used only preemptive analgesia [72, 74, 76] and one considered a second 12 h post-operative administration of the same pre-emptive drugs [69]. Totally, three different PGL pre-emptive dosages (7 mg, 10 mg and 300 mg) have been studied. Jokela et al. confronted different dosages (300 mg and 600 mg) of perioperative of PGL alone with diazepam 10 mg as control group after laparoscopic hysterectomy concluding that only PGL 600 mg, decreases oxycodone consumption postoperatively, and is associated with an increased incidence of dizziness, blurred vision, and headache [71]. However, when the same authors tested the pre-emptive PGL 10 or 7 mg in combination with ibuprofen in a similar setting, they did not found differences about neither the amount of postoperative analgesics required, nor the incidence of sideeffects [72]. No studies considered the comparison of preemptive and post-incisional or post-surgery PGL administration. The follow up period was no more than 24 h in 6 RCTs [69, 70, 72, 74, 7, 76], one trial followed the patients for 72 hours [71] and the the other once a day till 7 POD [73]. No studies established the optimal post surgical PGL treatment duration. Five studies assessed pain by using VAS [71, 72, 74-76], one study used both VAS and 4-point VRS [70], and the other two used an 11-point VRS [69, 73]. Six RCTs evaluated pain both at rest and on movement [69, 70, 72, 74-76] and the remaining 2 works did not specify whether pain was measured at rest or with movement [71, 73]. All the PGL-studies used the total rescue analgesics consumption as an important outcome for testing post-operative analgesia level. Morphine was delivered by the patients instructed to use an i.v. PCA pump in three RCTs [69, 74, 76], fentanyl [7] or oxycodone [71] in one trial respectively, without